Salvato in:
Dettagli Bibliografici
Autori principali: Song, Fanghui, Wang, Zhongjian, Sun, Jiebao
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2511.06042
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866908638909562880
author Song, Fanghui
Wang, Zhongjian
Sun, Jiebao
author_facet Song, Fanghui
Wang, Zhongjian
Sun, Jiebao
contents We propose a consistency model based on the optimal-transport flow. A physics-informed design of partially input-convex neural networks (PICNN) plays a central role in constructing the flow field that emulates the displacement interpolation. During the training stage, we couple the Hamilton-Jacobi (HJ) residual in the OT formulation with the original flow matching loss function. Our approach avoids inner optimization subproblems that are present in previous one-step OFM approaches. During the prediction stage, our approach supports both one-step (Brenier-map) and multi-step ODE sampling from the same learned potential, leveraging the straightness of the OT flow. We validate scalability and performance on standard OT benchmarks.
format Preprint
id arxiv_https___arxiv_org_abs_2511_06042
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Physics-Informed Design of Input Convex Neural Networks for Consistency Optimal Transport Flow Matching
Song, Fanghui
Wang, Zhongjian
Sun, Jiebao
Machine Learning
We propose a consistency model based on the optimal-transport flow. A physics-informed design of partially input-convex neural networks (PICNN) plays a central role in constructing the flow field that emulates the displacement interpolation. During the training stage, we couple the Hamilton-Jacobi (HJ) residual in the OT formulation with the original flow matching loss function. Our approach avoids inner optimization subproblems that are present in previous one-step OFM approaches. During the prediction stage, our approach supports both one-step (Brenier-map) and multi-step ODE sampling from the same learned potential, leveraging the straightness of the OT flow. We validate scalability and performance on standard OT benchmarks.
title Physics-Informed Design of Input Convex Neural Networks for Consistency Optimal Transport Flow Matching
topic Machine Learning
url https://arxiv.org/abs/2511.06042